This proposal is a straightforward, prospective randomized clinical trial designed to determine, in clinically node negative breast cancer patients undergoing breast conserving therapy, if sentinel node resection (and nodal irradiation to sentinel node positive patients) will provide, with considerable reduction in morbidity, the same regional control and prognostic information as axillary lymphadenectomy. Successful completion of this proposal has the potential to demonstrate that the majority of breast cancer patients can be spared the morbid and costly procedure of axillary lymphadenectomy. We hypothesize that 1) sentinel node biopsy (and nodal irradiation only if the sentinel node is pathologically positive) will result in regional control as effective as axillary lymphadenectomy, 2) staging and prognosis based on the sentinel node is accurate and equivalent to conventional axillary lymphadenectomy, 3) morbidity of sentinel node resection (and nodal irradiation to sentinel node positive patients) is less than conventional axillary lymphadenectomy, and 4) a group of patients at potentially increased risk of systemic recurrence can be identified which is sentinel node negative by conventional pathological analysis but positive by more detailed pathological investigation. Specific Aim 1: Determine that the rate of regional control achieved by sentinel node resection (and nodal irradiation only to those patients with pathologically positive sentinel nodes) is equivalent to conventional axillary lymphadenectomy. Specific Aim 2: In breast cancer patients with clinically node negative disease, determine the prognostic accuracy of the pathological status of the sentinel node. Specific Aim 3: Determine that the morbidity of sentinel node resection (with nodal irradiation to selected patients) is significantly less than conventional axillary lymphadenectomy in all patients. Specific Aim 4: Determine whether more intensive analysis of the sentinel lymph nodes results in more frequent identification of metastases and establish a data base relating this data to survival outcomes.